Prognosis evaluation and follow-up scheme generation method and device based on large model and medium

Abstract

The invention discloses a prognosis evaluation and follow-up scheme generation method and device based on a large model and a medium, and relates to the technical field of artificial intelligence. The method comprises the following steps: collecting multi-modal medical data corresponding to a patient, and preprocessing the multi-modal medical data to form standard medical data; wherein the multi-modal medical data comprises clinical data and image data; based on the standard medical data, performing retrieval in a preset standard knowledge base to obtain to-be-applied evaluation data; processing the to-be-applied evaluation data and the standard medical data by using a preset prognosis evaluation model to generate a corresponding prognosis evaluation index; and generating an optimal follow-up scheme based on the prognosis evaluation index and the to-be-applied evaluation data.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence technology, and in particular to a method, device and medium for generating a prognosis assessment and follow-up plan based on a large model. Background Art

[0002] With the improvement of global medical standards, personalized medicine and precision medicine have gradually become important development directions in the medical industry. The core of personalized medicine is to develop the most appropriate treatment and follow-up plan based on the individual differences of patients, including medical history, genetic data, living habits, etc., so as to maximize the treatment effect. However, existing medical follow-up plans often rely on the experience of doctors and traditional fixed follow-up cycles, and cannot be adjusted in real time according to the actual development of the patient's condition. This method is not only inefficient, but also due to the lack of personalized considerations, it is easy to lead to unsatisfactory treatment effects and even delay the patient's recovery process.

[0003] In actual practice, the sources of patients' follow-up data are wide and diverse, including medical records, laboratory test results, medical images, drug use and other multimodal medical data. These data are scattered and complex in format, making them difficult to integrate effectively. Traditional follow-up plan generation and adjustment methods lack the ability to deeply integrate and analyze these multimodal medical data, and usually rely only on single-dimensional data, ignoring the correlation between multi-dimensional information. This not only reduces the efficiency of data utilization, but also greatly reduces the accuracy of prognosis assessment, and cannot provide patients with truly accurate follow-up guidance.

[0004] In addition, the current prognosis assessment and follow-up plan generation system is often based on static data. Once the follow-up plan is formulated, it is difficult to make timely adjustments based on the patient's subsequent examination results, changes in condition, or treatment feedback. This lack of dynamic updates and closed-loop feedback mechanisms leads to a lag in follow-up results and is unable to keep pace with the patient's condition changes. Especially for patients with complex or rapidly changing conditions, the lag of traditional plans may bring serious medical risks.

[0005] The rise of big model technology has brought new opportunities for personalized medicine. Through the deep learning and analysis of medical knowledge bases and multimodal medical data by big models, more accurate prognosis assessment and personalized follow-up plan generation can be achieved. At the same time, the natural language processing and image processing capabilities of big models can effectively integrate text and image data to provide comprehensive diagnostic support for patients.

[0006] However, the current large model-driven medical system still faces technical challenges such as integrating multimodal medical data, improving real-time response capabilities, and dynamically adjusting follow-up plans. Therefore, there is an urgent need for an intelligent system that can integrate multimodal medical data and achieve closed-loop optimization based on large model technology to improve the dynamic adjustment capabilities of follow-up plans, improve patient treatment effects and compliance, and promote the advancement of personalized medicine. Summary of the invention

[0007] The embodiments of the present application provide a method, device and medium for prognosis assessment and follow-up plan generation based on a big model, which can integrate multimodal medical data and realize closed-loop optimization based on big model technology, so as to enhance the dynamic adjustment capability of the follow-up plan and improve the treatment effect and compliance of patients.

[0008] In a first aspect, an embodiment of the present application provides a method for generating a prognosis assessment and follow-up plan based on a large model, the method comprising: collecting multimodal medical data corresponding to the patient, and preprocessing the multimodal medical data to form standard medical data; wherein the multimodal medical data comprises: clinical data and imaging data; based on the standard medical data, searching in a preset standard knowledge base to obtain evaluation data to be applied; using a preset prognosis assessment model, processing the evaluation data to be applied and the standard medical data to generate corresponding prognosis assessment indicators; based on the prognosis assessment indicators and the evaluation data to be applied, generating an optimal follow-up plan.

[0009] In one implementation of the present application, before collecting multimodal medical data corresponding to the patient and preprocessing the multimodal medical data to form standard medical data, the method also includes: building a standard knowledge base based on a preset medical data source, specifically including: obtaining multimodal data in the medical data source, and preprocessing the multimodal data to obtain corresponding standard data; wherein, the standard data includes but is not limited to at least one of the following: literature guidelines, clinical cases; and storing the standard data in the standard knowledge base using a combination of a relational database and a vector database.

[0010] In one implementation of the present application, clinical data includes but is not limited to at least one of the following: medical records, examination results, and surgical records; imaging data includes but is not limited to at least one of the following: X-ray, CT, MRI; pre-processing multimodal medical data to form standard medical data, specifically including: integrating clinical data through a preset clinical data integration template to obtain standard clinical information; processing image data through a preset image analysis tool to extract key image features and image information.

[0011] In one implementation of the present application, the evaluation materials to be applied include: literature guidelines to be applied, clinical cases to be applied; based on standard medical data, a search is performed in a preset standard knowledge base to obtain the evaluation materials to be applied, specifically including: based on standard medical data, a search is performed in a relational database using a keyword matching method to obtain literature guidelines to be applied; and, based on standard medical data, a search is performed in a vector database using a similarity matching method to obtain clinical cases to be applied.

[0012] In one implementation of the present application, before using a preset prognosis evaluation model to process the application evaluation data and standard medical data, the method also includes: obtaining a prognosis evaluation data set, and training and annotating the prognosis evaluation data set; wherein the prognosis evaluation data set includes: standard medical data of historical patients, prognosis evaluation results and corresponding follow-up plan effect feedback; using the prognosis evaluation data set to train a deep learning model until a converged prognosis evaluation model is obtained; cross-validating the prognosis evaluation model, and adjusting the structure or parameters of the prognosis evaluation model according to the validation results until the preset evaluation accuracy requirements are met.

[0013] In one implementation of the present application, based on the prognostic evaluation index and the evaluation data to be applied, an optimal follow-up plan is generated, specifically including: inputting the prognostic evaluation index and the evaluation data to be applied into a preset optimal follow-up plan generation algorithm to generate an optimal follow-up plan; wherein the optimal follow-up plan generation algorithm is represented by the following formula:

[0014] P = argmax P (U(P|I,K)-C(P))

[0015] Among them, U(P|I,K) is the utility function of the follow-up plan, which is used to indicate the degree of improvement of the plan on the patient's health status, and C(P) is the implementation cost of the plan.

[0016] In one implementation of the present application, the method also includes: regularly collecting the patient's health status update data during the implementation of the follow-up plan; inputting the health status update data into the prognosis evaluation model to re-evaluate the patient's prognosis, and dynamically adjusting the follow-up plan based on the results of the re-evaluation.

[0017] In one implementation of the present application, the follow-up plan is dynamically adjusted according to the results of the re-evaluation, specifically including: analyzing the differences between the updated health status data and the standard medical data to identify the key indicators of changes in the patient's condition; based on the key indicators, re-searching the standard knowledge base to obtain updated evaluation data; using the updated evaluation data and the updated health status data, re-running the prognosis evaluation model to generate updated prognosis evaluation indicators; based on the updated prognosis evaluation indicators and the updated evaluation data, using the optimal follow-up plan generation algorithm, calculating the adjusted follow-up plan, and promptly notifying relevant personnel to execute it.

[0018] In a second aspect, an embodiment of the present application also provides a large-model-based prognosis assessment and follow-up plan generation device, characterized in that the device includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a large-model-based prognosis assessment and follow-up plan generation method such as any one of the above.

[0019] In a third aspect, the present application also provides a non-volatile computer storage medium for generating a prognosis assessment and follow-up plan based on a large model, which stores computer executable instructions, characterized in that when the computer executable instructions are executed, a method for generating a prognosis assessment and follow-up plan based on a large model as described above is implemented.

[0020] The embodiments of the present application provide a method, device and medium for generating a prognosis assessment and follow-up plan based on a large model, which has the following beneficial effects:

[0021] 1. By introducing large model technology, this application can achieve deep learning and analysis of multimodal medical data of patients, and generate more accurate prognosis evaluation indicators by combining with the preset standard knowledge base. Compared with traditional methods, this application significantly improves the accuracy of prognosis evaluation and provides doctors with more reliable decision support.

[0022] 2. Based on the prognostic evaluation indicators and the patient's individualized information, this application can generate the optimal follow-up plan to ensure that the follow-up plan is more in line with the patient's actual needs and changes in the condition, which helps to improve the treatment effect and accelerate the patient's recovery process.

[0023] 3. Through automated processing and dynamic adjustment mechanisms, this application reduces the manual operations of doctors during follow-up and improves follow-up efficiency. At the same time, the real-time updated follow-up plan enables doctors to respond more quickly to changes in patients' conditions and provide patients with more timely and effective treatment.

[0024] 4. Through accurate prognostic assessment and personalized follow-up plans, this application helps to optimize the allocation of medical resources. Doctors can arrange follow-up time and resources more reasonably, reduce unnecessary waste of medical resources, and improve the overall efficiency and quality of medical services. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0026] Figure 1 A flow chart of a method for generating a prognosis assessment and follow-up plan based on a large model provided in an embodiment of the present application;

[0027] Figure 2 A schematic diagram of the internal structure of a large-model-based prognosis assessment and follow-up plan generation device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0028] In order to make the purpose, technical solution and advantages of the present application clearer, the technical solution of the present application will be clearly and completely described below in combination with the specific embodiments of the present application and the corresponding drawings. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present application.

[0029] The embodiments of the present application provide a method, device and medium for prognosis assessment and follow-up plan generation based on a big model, which can integrate multimodal medical data and realize closed-loop optimization based on big model technology, so as to enhance the dynamic adjustment capability of the follow-up plan and improve the treatment effect and compliance of patients.

[0030] The technical solution proposed in the embodiments of the present application is described in detail below with reference to the accompanying drawings.

[0031] Figure 1 A flow chart of a method for generating a prognosis assessment and follow-up plan based on a large model provided in an embodiment of the present application. Figure 1 As shown, the embodiment of the present application provides a method for generating a prognosis assessment and follow-up plan based on a large model, which specifically includes the following steps:

[0032] Step 101: Collect multimodal medical data corresponding to the patient, and pre-process the multimodal medical data to form standard medical data.

[0033] In one embodiment of the present application, in order to achieve prognosis assessment and follow-up plan generation, it is first necessary to build a standard knowledge base based on a preset medical data source.

[0034] Specifically, multimodal data in a medical data source are obtained, and the multimodal data are preprocessed to obtain corresponding standard data; wherein the standard data includes but is not limited to at least one of the following: literature guidelines, clinical cases; and the standard data are stored in a standard knowledge base by combining a relational database and a vector database.

[0035] In one embodiment, first, multimodal data are collected from multiple medical data sources such as public medical databases, clinical guidelines, and drug databases. These data include, but are not limited to, literature guidelines, clinical cases, etc., which cover a wealth of medical knowledge and practical experience. The collected multimodal data need to be preprocessed to form unified standard data. The preprocessing process may include steps such as data cleaning, format conversion, and information extraction to ensure the accuracy and consistency of the data. The preprocessed multimodal data are organized into standard data; it is understandable that these standard data not only include the original literature guidelines and clinical case content, but also may include key information extracted from the data, structured data, etc., for subsequent retrieval and analysis. In order to efficiently manage and retrieve standard data, this application adopts a combination of relational databases and vector databases. The relational database is used to store structured data, such as authors, publication years, keywords, etc. in the literature guidelines; and the vector database is used to store unstructured data, such as descriptions of clinical cases, image data, etc., and efficient retrieval is achieved through vectorization. The combined use of the two databases gives full play to their respective advantages and improves the accuracy and efficiency of data retrieval.

[0036] In one embodiment of the present application, after building a standard knowledge base, if prognosis evaluation and follow-up plan generation are required, it is first necessary to collect the multimodal medical data corresponding to the patient and pre-process the multimodal medical data to form standard medical data. It should be noted that the multimodal medical data in the embodiment of the present application includes: clinical data and imaging data; clinical data includes but is not limited to medical records, examination results, surgical records, etc., and imaging data includes but is not limited to: X-ray, CT, MRI, etc.

[0037] Preprocessing multimodal medical data to form standard medical data, specifically including: integrating clinical data through preset clinical data integration templates to obtain standard clinical information; processing image data through preset image analysis tools to extract key image features and image information.

[0038] In one embodiment, the preset clinical data integration template includes preset fields and formats for unifying clinical data from different sources. During the integration process, text data needs to be converted into structured data, such as converting free text in medical records into standardized diagnostic codes or symptom descriptions. The integrated clinical data forms standard clinical information, providing a clear and consistent data basis for subsequent analysis. Image analysis tools include algorithms such as image segmentation and feature extraction, which are used to extract key image features and image information from image data. For example, the lesion area can be identified by an image segmentation algorithm, and the shape, size, density and other features of the lesion can be extracted by a feature extraction algorithm. The processed image data not only contains the original image information, but also contains rich quantitative features, which provides strong support for the generation of prognosis evaluation and follow-up plans.

[0039] Step 102: Based on the standard medical data, a search is performed in a preset standard knowledge base to obtain evaluation data to be applied.

[0040] In one embodiment of the present application, after preprocessing the multimodal medical data to form standard medical data, a search is performed in a preset standard knowledge base based on the standard medical data to obtain evaluation materials to be applied, wherein the evaluation materials to be applied include: literature guidelines to be applied and clinical cases to be applied.

[0041] Specifically, based on standard medical data, a keyword matching method is used to search in a relational database to obtain literature guidelines to be applied; and based on standard medical data, a similarity matching method is used to search in a vector database to obtain clinical cases to be applied.

[0042] In one embodiment, in order to obtain literature guides related to standard medical data, this application uses a keyword matching method to search in a relational database. First, extract keywords from the standard medical data. These keywords may include disease names, symptom descriptions, test results, etc. Then, search for literature guide records containing these keywords in the relational database. The database management system will sort the search results according to the keyword matching degree, and give priority to returning the literature guides with the highest matching degree. In this way, the literature guides most relevant to the standard medical data can be quickly located, providing a theoretical basis for prognosis evaluation.

[0043] In order to obtain clinical cases similar to standard medical data, this application uses a similarity matching method to search in a vector database. First, the standard medical data is converted into a vector representation, which usually involves extracting key features and vectorizing them; then, the vector database is searched for clinical case vectors that are similar to the standard medical data vectors. The similarity matching method can evaluate the similarity by calculating indicators such as cosine similarity and Euclidean distance between vectors. The database management system will sort the search results according to the similarity score and give priority to returning clinical cases with the highest similarity. These clinical cases will provide practical experience references for the generation of follow-up plans.

[0044] After obtaining the literature guidelines and clinical cases to be applied, comprehensive evaluation and screening are required, including the evaluation of the relevance, reliability, and timeliness of the search results, to ensure that the selected data can provide valuable support for prognosis evaluation and the generation of follow-up plans. At the same time, according to the specific situation and needs of the patient, the search results are screened individually to obtain the most suitable evaluation data for the patient.

[0045] Step 103: Using a preset prognostic evaluation model, the application evaluation data and standard medical data are processed to generate corresponding prognostic evaluation indicators.

[0046] In one embodiment of the present application, after obtaining the assessment data to be applied, the assessment data to be applied and standard medical data are processed using a preset prognostic assessment model to generate corresponding prognostic assessment indicators.

[0047] In one embodiment of the present application, the construction of the prognosis evaluation model specifically includes the following processes: obtaining a prognosis evaluation data set, and training and annotating the prognosis evaluation data set; wherein the prognosis evaluation data set includes: standard medical data of historical patients, prognosis evaluation results and corresponding follow-up program effect feedback; using the prognosis evaluation data set to train the deep learning model until a converged prognosis evaluation model is obtained; cross-validating the prognosis evaluation model, and adjusting the structure or parameters of the prognosis evaluation model according to the validation results until the preset evaluation accuracy requirements are met.

[0048] Step 104: Generate an optimal follow-up plan based on the prognostic evaluation indicators and the evaluation data to be applied.

[0049] In one embodiment of the present application, after the corresponding prognostic evaluation index is generated, an optimal follow-up plan is generated based on the prognostic evaluation index and the evaluation data to be applied.

[0050] Specifically, the prognostic evaluation index and the evaluation data to be applied are input into a preset optimal follow-up plan generation algorithm to generate an optimal follow-up plan; wherein the optimal follow-up plan generation algorithm is represented by the following formula:

[0051] P = argmax P (U(P|I,K)-C(P))

[0052] Among them, U(P|I,K) is the utility function of the follow-up plan, which is used to indicate the degree of improvement of the plan on the patient's health status, and C(P) is the implementation cost of the plan.

[0053] In one embodiment of the present application, the method also includes: regularly collecting the patient's health status update data during the implementation of the follow-up plan; inputting the health status update data into the prognosis assessment model to re-evaluate the patient's prognosis, and dynamically adjusting the follow-up plan based on the results of the re-evaluation.

[0054] In one embodiment of the present application, the follow-up plan is dynamically adjusted according to the results of the re-evaluation, specifically including: analyzing the differences between the updated health status data and the standard medical data to identify the key indicators of changes in the patient's condition; based on the key indicators, re-searching the standard knowledge base to obtain updated evaluation data; using the updated evaluation data and the updated health status data, re-running the prognosis evaluation model to generate updated prognosis evaluation indicators; based on the updated prognosis evaluation indicators and the updated evaluation data, using the optimal follow-up plan generation algorithm, calculating the adjusted follow-up plan, and promptly notifying relevant personnel to execute it.

[0055] The above is an embodiment of the method proposed in this application. Based on the same inventive concept, the embodiment of this application also provides a large model-based prognosis assessment and follow-up plan generation device, the structure of which is as follows: Figure 2 shown.

[0056] Figure 2 A schematic diagram of the internal structure of a large model-based prognosis assessment and follow-up plan generation device provided in an embodiment of the present application. Figure 2 As shown, the device includes:

[0057] at least one processor 201;

[0058] and, a memory 202 communicatively connected to the at least one processor;

[0059] The memory 202 stores instructions that can be executed by at least one processor, and the instructions are executed by at least one processor 201 to enable at least one processor 201 to:

[0060] Collecting multimodal medical data corresponding to the patient and preprocessing the multimodal medical data to form standard medical data; wherein the multimodal medical data includes: clinical data and imaging data;

[0061] Based on standard medical data, search in the preset standard knowledge base to obtain the evaluation data to be applied;

[0062] Using the preset prognostic evaluation model, the application evaluation data and standard medical data are processed to generate corresponding prognostic evaluation indicators;

[0063] Generate the optimal follow-up plan based on the prognostic evaluation indicators and the evaluation data to be applied.

[0064] Some embodiments of the present application provide corresponding Figure 1 A non-volatile computer storage medium for generating a prognosis evaluation and follow-up plan based on a large model, storing computer executable instructions, wherein the computer executable instructions are set to:

[0065] Collecting multimodal medical data corresponding to the patient and preprocessing the multimodal medical data to form standard medical data; wherein the multimodal medical data includes: clinical data and imaging data;

[0066] Based on standard medical data, search in the preset standard knowledge base to obtain the evaluation data to be applied;

[0067] Using the preset prognostic evaluation model, the application evaluation data and standard medical data are processed to generate corresponding prognostic evaluation indicators;

[0068] Generate the optimal follow-up plan based on the prognostic evaluation indicators and the evaluation data to be applied.

[0069] Each embodiment in this application is described in a progressive manner, and the same or similar parts between the embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the IoT device and medium embodiments, since they are basically similar to the method embodiments, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiments.

[0070] The system and medium provided in the embodiments of the present application correspond one-to-one to the method. Therefore, the system and medium also have similar beneficial technical effects to the corresponding methods. Since the beneficial technical effects of the methods have been described in detail above, the beneficial technical effects of the system and medium will not be repeated here.

[0071] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0072] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0073] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0074] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0075] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0076] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0077] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0078] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

[0079] The above is only an embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application should be included in the scope of the claims of the present application.

Claims

1. A method for prognosis assessment and follow-up plan generation based on a large model, characterized in that: The method comprises: Collecting multimodal medical data corresponding to the patient, and preprocessing the multimodal medical data to form standard medical data; wherein the multimodal medical data includes: clinical data and imaging data; Based on the standard medical data, searching in a preset standard knowledge base to obtain evaluation data to be applied; Using a preset prognostic evaluation model, the evaluation data to be applied and the standard medical data are processed to generate corresponding prognostic evaluation indicators; Based on the prognostic evaluation indicators and the evaluation data to be applied, an optimal follow-up plan is generated.

2. A method for generating a prognosis assessment and follow-up plan based on a large model according to claim 1, characterized in that: Before collecting the multimodal medical data corresponding to the patient and preprocessing the multimodal medical data to form standard medical data, the method further includes: Based on the preset medical data sources, a standard knowledge base is built, including: Acquire multimodal data from the medical data source, and preprocess the multimodal data to obtain corresponding standard data; wherein the standard data includes but is not limited to at least one of the following: literature guidelines, clinical cases; The standard data is stored in the standard knowledge base by combining a relational database and a vector database.

3. A method for generating a prognosis assessment and follow-up plan based on a large model according to claim 2, characterized in that: The clinical data include but are not limited to at least one of the following: medical records, examination results, and surgical records; the imaging data include but are not limited to at least one of the following: X-ray, CT, and MRI; Preprocessing the multimodal medical data to form standard medical data specifically includes: Integrate the clinical data using a preset clinical data integration template to obtain standard clinical information; The image data is processed by a preset image analysis tool to extract key image features and image information.

4. A method for generating a prognosis assessment and follow-up plan based on a large model according to claim 3, characterized in that: The evaluation materials to be applied include: literature guidelines to be applied, clinical cases to be applied; Based on the standard medical data, a search is performed in a preset standard knowledge base to obtain evaluation data to be applied, specifically including: Based on the standard medical data, searching in a relational database using a keyword matching method to obtain literature guidelines to be applied; and, Based on the standard medical data, a similarity matching method is used to search in a vector database to obtain clinical cases to be applied.

5. The method for generating a prognosis assessment and follow-up plan based on a large model according to claim 1, characterized in that: Before processing the evaluation data to be applied and the standard medical data using the preset prognostic evaluation model, the method further includes: Obtaining a prognosis assessment data set, and training and annotating the prognosis assessment data set; wherein the prognosis assessment data set includes: standard medical data of historical patients, prognosis assessment results, and corresponding follow-up program effect feedback; Using the prognostic assessment data set to train a deep learning model until a converged prognostic assessment model is obtained; The prognosis evaluation model is cross-validated, and the structure or parameters of the prognosis evaluation model are adjusted according to the validation results until the preset evaluation accuracy requirement is met.

6. The method for generating a prognosis assessment and follow-up plan based on a large model according to claim 1, characterized in that: Based on the prognostic evaluation index and the evaluation data to be applied, an optimal follow-up plan is generated, which specifically includes: Inputting the prognostic evaluation index and the evaluation data to be applied into a preset optimal follow-up plan generation algorithm to generate an optimal follow-up plan; The optimal follow-up plan generation algorithm is expressed by the following formula: P=argmax P (U(P|I,K)-C(P)) Among them, U(P|I,K) is the utility function of the follow-up plan, which is used to indicate the degree of improvement of the plan on the patient's health status, and C(P) is the implementation cost of the plan.

7. The method for generating a prognosis assessment and follow-up plan based on a large model according to claim 1, characterized in that: The method further comprises: During the follow-up protocol, data updates on the patient's health status are collected regularly; The updated health status data is input into the prognosis assessment model to reassess the patient's prognosis, and the follow-up plan is dynamically adjusted according to the results of the reassessment.

8. A method for generating a prognosis assessment and follow-up plan based on a large model according to claim 7, characterized in that: According to the results of the re-evaluation, the follow-up plan is adjusted dynamically, including: Analyze the difference between the health status update data and the standard medical data to identify key indicators of changes in the patient's condition; Based on the key indicators, re-search the standard knowledge base to obtain updated evaluation information; Re-running the prognostic assessment model using the updated assessment information and the updated health status data to generate an updated prognostic assessment indicator; According to the updated prognostic evaluation indicators and the updated evaluation data, an optimal follow-up plan generation algorithm is used to calculate the adjusted follow-up plan, and relevant personnel are notified in a timely manner to implement it.

9. A device for generating prognosis assessment and follow-up plans based on a large model, characterized in that: The device comprises: at least one processor; and, a memory communicatively coupled to the at least one processor; Wherein, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute a large model-based prognosis assessment and follow-up plan generation method as described in any one of claims 1-8.

10. A non-volatile computer storage medium for generating a prognosis assessment and follow-up plan based on a large model, storing computer executable instructions, characterized in that: When the computer executable instructions are executed, a large model-based prognosis assessment and follow-up plan generation method as described in any one of claims 1 to 8 is implemented.

Images